Home Case Study Transforming Government through AI and Data Science

Transforming Government through AI and Data Science

0

Data Science Clouds

With a growing investment into local skills through large-scale funding and courses for students, ‘Data Science’ is a powerful capability for the Scottish public sector to harness, especially so when combined with other technologies like Cloud computing.

The tier one providers like Microsoft Azure offer a suite of integrated Cloud services to this end. As their European CTO Nigel Willson tweets:

Service options include the Data Catalog providing all the tools required for data scientists to build and analyze complex data sets, such as PowerBI and HDInsight to manage, analyze, and visualize large quantities of data, and an AI suite which enables building custom models with Azure Machine Learning for any scenario.

Microsoft also recently open sourcing an AI engine, and Data Scientists can also leverage their R programming skills on Azure.

From Reaction to Prevention

It’s great that these are currently very popular topics but like any technologies it’s essential that first they need a clearly defined purpose for their implementation.

There’s a litany of stories of government’s pouring £ millions into the latest and greatest technologies only for them to fail to produce any meaningful outcome for citizens, so the first critical step is to articulate a clear vision for the transformation the investment will realize.

For government this purpose can be very succinctly articulated through the simplest of ideals, a shift from a preventative rather than reactive approach to social and healthcare delivery, a transition that can simultaneously improve services while also reducing costs, so an entirely compelling ROI business case.

A number of case studies illustrate this principle in action.

The first for Boulder County documents their adoption of an integrated service delivery system, guided by a simple but powerful objective at the heart of their transformation:

focus on front-end and early intervention measures to prevent more costly services in the future.

It sounds an obvious goal but the reality for most stretched case workers is they are always in a reactive mode, dealing with the consequences of social challenges. These consequences have multiple levels of cascading impacts and costs for government, and the use of data science yields insights which instead enables them to proactively tackle those issues before the impacts occur.

They offer a number of lessons learned that are especially relevant to Scotland’s goals. Notably the ambition to realize an integrated approach to Health and Social care. Boulder County had also previously operated distinct departments for each service, and:

in 2008 the County began a system-wide shift to co-create solutions for complex family and community challenges by fully integrating health, housing, and human services. The idea was to generate a more self-sufficient, sustainable, and resilient community by focusing on reducing the social determinants of poor health, removing barriers to services, and moving the system upstream towards an early intervention and prevention model.

They also incorporated and integrated services from the Third Sector, and by pooling information from across previously siloed agencies and applying Data Science they were able to shift on to the front foot of more upstream prevention rather than downstream reaction:

They had come to understand that by having a comprehensive view of each client’s situation, caseworkers are better able to identify opportunities to apply the early intervention and prevention approach to wrap-around services and help the client stabilize.

“We know that 70 percent of our Section 8 (housing choice voucher) clients are also receiving food assistance,” one caseworker said. “So how can we work together with the housing/Section 8 and the food case managers to improve services?” To further broaden the reach of the integrated case management system, county attorneys drafted memoranda of understanding to incorporate community partners such as nonprofits and community health centers.

Crime is a key scenario for illustrating this effect. In Predictive Tools for Public Safety the author describes a number of case studies of the police using data analytics to proactively deter crime, rather than always be responding to crimes after they happen. This can eliminate multiple costs such as the damages caused, prison and court system expenses.

Santa Cruz turned to an applied mathematician to develop PredPol, for Predictive Policing, which analyzes previous property crimes to predict where future ones will occur, and plots these hot spots on maps, which are provided to officers at shift briefings for them to utilize for purposes of targeted patrol to prevent the crimes.

In 2006 violent re-offenders established Philadelphia as one of the murder capitals of the USA, and to tackle the issue they employed machine learning algorithms for probationer backgrounds to estimate the likelihood of violent re-offense, prioritizing resources accordingly to reduce the likelihood of future crimes.

Data-driven Strategies for Reducing Homelessness describes how New York City used analytics to proactively identify which families most likely to face homelessness, and intervene to prevent this occurring. HomeBase reduced the number of shelter applications by nearly 50% and reduced the number of days in shelter by 70%, resulting in $1.37 in savings for every dollar spent on the program.

The Cloud Best Practices Network is a global practitioner community for Digital Transformation and Cloud Computing.

LEAVE YOUR COMMENT

Your email address will not be published. Required fields are marked *